How NVIDIA And Insilico Medicine Are Accelerating Drug Discovery With Generative AI
27 June 2024
Businesses often find themselves at a crossroads in the race to leverage artificial intelligence (AI). The lure of AI’s promise is undeniable—from enhancing customer experiences to automating routine tasks. Yet, how a company approaches AI can spell the difference between mere technological flirtation and achieving real, transformative outcomes. Here, I offer a strategic blueprint for businesses keen on not just piloting AI but also scaling it effectively.

Generative AI is the technology behind ChatGPT and other conversational, content-generating tools. And while the technology can be used in all sorts of fun and interesting ways, it also has a very serious role to play in the world of drug discovery. Read on to discover how generative AI is helping pharma companies develop new drugs, faster.
The Role Of Generative AI In Drug Discovery
I had the pleasure of interviewing Kimberly Powell, VP and general manager of healthcare at NVIDIA, for my podcast. Powell’s team works with industry leaders, academics, pharma companies, and biotech companies to apply AI to the drug discovery process. So, when it comes to figuring out how generative AI can aid the search for new drug therapies, Powell is the person to ask.
We tend to think of generative AI in terms of understanding and generating human language. But what gets the team at NVIDIA really excited is the potential to apply generative AI to other languages related to drug development: the languages of biology and chemistry. Think of human DNA as a sequence of four letters (A, T, C and G) strung together into a 3-billion-letter long sentence. That's a language of its own. We also have proteins, which form the building blocks of biology. Proteins have their own alphabet – 20 letters for amino acids, strung together in lengths of tens of thousands or even hundreds of thousands. Chemicals, too, have a language known as SMILES (Simplified Molecular Input Line Entry System) – characters that, together, define the structures of chemistry.
As Powell puts it, "We can now take these languages … and we can apply the method of generative AI and GPT-type methods … Once we do that, the language models can really help us understand a lot more about biology that maybe we haven't been able to observe in the real world." This means that not only can we discover new drugs with generative AI, but we can also do it in less time and at a lower cost. Considering that the failure rate for new drug therapies is 90 percent – meaning only 10 percent end up making it through clinical trials and being approved for clinical use – anything that can speed up and improve the drug discovery process could have a huge impact.
Adding Value At Each Stage Of The Process
But how exactly does generative AI apply to the drug discovery process? Powell described the drug discovery process as having three phases:
The first phase is about establishing the target (disease or condition) that they want to treat with a new drug. In this phase, generative AI can be used to study genomics, understand the gene that’s causing the disease, or understand whatever is happening in the body. Basically, to understand the target better.
The second phase of drug discovery is about coming up with leads, i.e., chemicals or proteins that could be used to target that disease. This is where the scale of the problem becomes truly mind-boggling because there are more than 1060 chemicals and 10160 proteins that could potentially be used to target a disease. No wonder drug discovery is often described as searching for a needle in a haystack! Generative AI can sift through these potential chemicals and proteins and start generating ideas – potentially even inventing new chemicals and proteins with the desired structure and function to target the disease in question. This creates a tremendous number of new leads to explore, which is really exciting.
And the third phase is about optimization. Say the generative model has generated one billion compounds that could potentially be effective, the drug company then needs to test those against the target. Generative AI can assist with this screening process at a scale and speed that’s never been seen before. In one example, NVIDIA worked on a project with Recursion Pharmaceuticals to screen more than 2.8 quadrillion small molecule-target pairs. Within a week, they were able to complete screening that would have taken 100,000 years with traditional methods.
In short, generative AI can help pharma companies explore potential new drugs with unprecedented scale, speed, and accuracy – which, in turn, allows them to proceed to clinical trials quicker.
How Insilico Medicine Deployed Generative AI
One inspiring example comes from Insilico Medicine, a biotech company and member of NVIDIA’s Inception program, which supplies cutting-edge AI training and support. Insilico used generative AI methods and NVIDIA technology to develop a drug to treat idiopathic pulmonary fibrosis – a relatively rare disease that causes progressive decline in lung function.
Using traditional methods, this process would have cost more than $400 million and taken up to six years. But thanks to generative AI, NVIDIA says Insilico accomplished the task for one-tenth of the cost and one-third of the time, proceeding to clinical trials in just two and a half years. What we’re talking about, then, is quicker cures for diseases. And for a fraction of the cost.
But how exactly did Insilico use generative AI to develop this drug? NVIDIA says generative AI was used at each step of the preclinical drug discovery process. Firstly, it was used to identify a specific molecule that a drug compound could target. It was then used to generate new drug candidates, and assess how well those novel drug candidates would bind with the target molecule. It was even used to predict the outcome of clinical trials.
Incidentally, Insilico has also made the news for developing a new AI-generated COVID drug that has entered clinical trials, and is reported to be effective against all variants. Plus, the company has more than 30 programs in the pipeline to target other diseases, including cancer – hopefully resulting in more success stories in the coming years.
Imagine this level of progress across all sorts of diseases and conditions, and it’s easy to see how generative AI is an absolute game-changer for the healthcare and pharmaceutical sectors. We can, therefore, expect generative AI to play a vital role in developing the drugs of the future.
Related Articles
7 Game-Changing Legal Tech Trends That Will Transform Law Firms In 2025
By now, “smart” versions exist of just about every home appliance, gadget and gizmos we can think of. However, manufacturers continue[...]
8 Critical Smart City Trends Reshaping Urban Life In 2025
By now, “smart” versions exist of just about every home appliance, gadget and gizmos we can think of. However, manufacturers continue[...]
The 8 Critical Energy Trends Defining 2025
By now, “smart” versions exist of just about every home appliance, gadget and gizmos we can think of. However, manufacturers continue[...]
16 Essential Generative AI Tools Transforming HR in 2025
By now, “smart” versions exist of just about every home appliance, gadget and gizmos we can think of. However, manufacturers continue[...]
Will AI Make Universal Basic Income Inevitable?
By now, “smart” versions exist of just about every home appliance, gadget and gizmos we can think of. However, manufacturers continue[...]
How to Spot AI Deepfake Videos In An Era Of Digital Deception
By now, “smart” versions exist of just about every home appliance, gadget and gizmos we can think of. However, manufacturers continue[...]
Sign up to Stay in Touch!
Bernard Marr is a world-renowned futurist, influencer and thought leader in the fields of business and technology, with a passion for using technology for the good of humanity.
He is a best-selling author of over 20 books, writes a regular column for Forbes and advises and coaches many of the world’s best-known organisations.
He has a combined following of 4 million people across his social media channels and newsletters and was ranked by LinkedIn as one of the top 5 business influencers in the world.
Bernard’s latest book is ‘Generative AI in Practice’.
Social Media